lstm-price-prediction-guide

LSTM Price Prediction Guide

A guide to building LSTM (Long Short-Term Memory) neural networks for cryptocurrency price prediction. Covers the full pipeline from data collection to model deployment.

Pipeline Overview

Data Collection → Feature Engineering → Preprocessing → Model Training → Evaluation → Deployment
     │                    │                  │                │               │            │
 CoinGecko,         Indicators,         Normalize,       LSTM layers,    RMSE/MAE,    REST API
 Binance API      Sentiment, Volume     Sequence data    Dropout, Dense   Backtesting   or MCP

Data Collection

import pandas as pd
from cryptodatapy import DataRequest

# Fetch historical OHLCV data
dr = DataRequest(
    tickers=['SPA'],
    fields=['open', 'high', 'low', 'close', 'volume'],
    freq='1h',
    start_date='2023-01-01',
    source='binance'
)
data = dr.fetch()

Feature Engineering

Technical Indicators

import ta

def add_features(df):
    # Moving averages
    df['sma_7'] = ta.trend.sma_indicator(df['close'], window=7)
    df['sma_25'] = ta.trend.sma_indicator(df['close'], window=25)
    df['ema_12'] = ta.trend.ema_indicator(df['close'], window=12)
    
    # RSI
    df['rsi'] = ta.momentum.rsi(df['close'], window=14)
    
    # MACD
    macd = ta.trend.MACD(df['close'])
    df['macd'] = macd.macd()
    df['macd_signal'] = macd.macd_signal()
    
    # Bollinger Bands
    bb = ta.volatility.BollingerBands(df['close'])
    df['bb_upper'] = bb.bollinger_hband()
    df['bb_lower'] = bb.bollinger_lband()
    
    # Volume indicators
    df['volume_sma'] = df['volume'].rolling(window=20).mean()
    df['volume_ratio'] = df['volume'] / df['volume_sma']
    
    # Returns
    df['returns'] = df['close'].pct_change()
    df['log_returns'] = np.log(df['close'] / df['close'].shift(1))
    
    return df.dropna()

Feature List

Feature	Type	Description
SMA (7, 25, 50)	Trend	Simple moving averages
EMA (12, 26)	Trend	Exponential moving averages
RSI (14)	Momentum	Relative strength index
MACD	Momentum	Moving average convergence
Bollinger Bands	Volatility	Price channels
Volume Ratio	Volume	Relative volume
Returns	Price	Percentage returns
ATR	Volatility	Average true range